Specific and Transfer Effects Induced by Arm or Leg ...

Training and Testing

Specific and Transfer Effects Induced by Arm or Leg Training N. Tordi1, A. Belli1, F. Mougin1, J. D. Rouillon1, M. Gimenez2 1

Laboratoire des Sciences du Sport, Place Saint Jacques, Besanc Ëon cedex ,France 2 Â 14 INSERM, C.O. 10, Vandoeuvre ls Nancy, France Unite

Tordi N, Belli A, Mougin F, Rouillon JD, Gimenez M. Specific and Transfer Effects Induced by Arm or Leg Training. Int J Sports Med 2001; 22: 517 ± 524 Accepted after revision: December 10, 2000

nnnn The purpose of the present study was to examine both the specific and the transfer effects of a Square Wave Endurance Exercise Test (SWEET) and to investigate the determining factors of transfer effect. A control group (CG, n = 5) and 2 experimental groups were studied. Experimental groups completed 3 days/ week a 45 min SWEET over 6 weeks, either with the arms using a wheelchair ergometer (AG, n = 5) or with the legs using a cycling ergometer (LG, n = 5). All subjects performed before and after training two maximal progressive tests: the first one with the arms and the second with the legs. During those tests oxyÇ O2) and cardiorespiratory parameters were congen uptake (V tinuously measured. Specific effects (changes between the tests performed with trained limbs) were observed: the peak power Ç O2 were significantly increased by both arm Ç p) and V output (W (+ 66 %, + 35 %) and leg (+ 17 %, + 14 %) training. At ventilatory Ç ) and V Ç O2 were significantly inthreshold (VT), power output (W creased with arm (+ 145 % and + 51 %) and leg (+ 53 % and + 46 %) training. Transfer effects (changes between pre and post tests Ç p and V Ç O2 performed with untrained limbs) were observed: W Ç p was increased in LG (+ 11 %). were not increased in AG while W Ç and V ÇO2 were increased with arm (+ 19 % and + 23 %) At VT, W Ç O2 and leg (+ 33 % and + 46 %) training. For AG and LG the V Ç O2 peak was increased (+ 19 % and expressed in percent of the V + 33 %, respectively) and the O2p was also increased (+ 30 %) for LG. These results show that SWEET training induced specific and transfer effects. Moreover, the initial level of the subjects, the type and the intensity of the training seem to be the major factors for effective observable transfer effects.

Introduction Regular endurance training enhances overall cardiorespiratory characteristics [2] and induces peripheral effects [7,10,12, 23, 26, 29, 33]. These adaptations generally increase the specific performance of the subjects (the performance measured during exercise with trained muscles). Transfer effects of endurance training (i. e., increase of performance measured during exercise with untrained muscles) is more controversial [9] than specific training effects. Some studies have reported that endurance training only affects the trained muscle groups [3,19, 22, 23, 29, 30, 36], whereas others have suggested that endurance training also increases the performance capacity of untrained muscles [4, 20, 21, 25, 31].

n Key words: Arm versus leg training, oxygen uptake, ventilatory threshold, interval training, wheelchair ergometer, SWEET.

A review of previous studies suggests that the occurrence of a transfer effect depends on the magnitude of the specific effect and that the transfer from legs to arms is more effective than transfer from arms to legs. When legs are trained, a 15 % increase in VÇO2p (measured in leg test) is generally sufficient to induce a significant transfer effect [4, 20, 31]. Only Stamford et al. [36] have not observed any transfer in arm exercise after an increase of 15 % of VÇO2p induced by a very specific leg training (bench stepping). In arm training studies, where the VÇO2p improvement (measured in arm test) is below 20 % [3, 22, 30, 36], no transfer effects are observed. An increase of more than 30 % of VÇO2p is necessary to obtain a significant transfer effect [20, 25]. It is recognized that the magnitude of specific effects depends on the pre-training level of VÇO2p of the subjects. It seems that initial VÇO2p also indirectly influences the transfer effect. A transfer effect of endurance training to untrained limbs is reported with subjects having a low VÇO2p [20, 21, 25]. These observations are consistent with the central circulatory changes observed after short periods of endurance training in individuals with low VÇO2p [38]. On the contrary, in studies with subjects having a relatively high initial VÇO2p no transfer effects were reported [3, 22, 30, 36]. Therefore, initial VÇO2p should be taken into consideration in studies dealing with transfer effect.

Int J Sports Med 2001; 22: 517 ± 524 Georg Thieme Verlag Stuttgart ´ New York ISSN 0172-4622

Endurance training also induces a better muscle oxidative capacity. The lactate threshold has been considered as a good indicator of muscle oxidative capacity [32] and endurance performance [35]. Using the ventilatory threshold (VT) in order to indirectly measure the lactate threshold from changes in the respiratory gas exchange variables during graded exercise

517

518

Int J Sports Med 2001; 22

Tordi N et al

[39], researchers [6, 8,11,16] have also shown that endurance training can significantly increase VÇO2 at VT. To the best of our knowledge, only two studies [3, 31] proposed to examine the transfer of training effects at the lactate threshold (LT). Bhambhani et al. [3] have shown any transfer effects on VÇO2 measured at VT after constant training with arms or legs. Rosler et al. [31] did not observe any changes in the power output at the level of 4 mmol l±1 lactate during tests with the untrained arms.

fore each test using gases of known concentration. All ventilatory variables were measured and averaged every thirty seconds. The respiratory and metabolic variables were oxygen uptake (VÇO2, ml min±1), CO2 production (VÇCO2, ml min±1), ventilation (VÇE, l min±1) and respiratory exchange ratio (RER = VÇO2/VÇCO2). Heart rate (HR, b min±1) was also monitored using a specific recording system (PE 4000 Sport Tester). Oxygen pulse (O2p = VÇO2/HR) was also calculated.

To sum up, transfer effect mainly depends on the magnitude of the specific effect (which is influenced by the initial level of subjects and by the type of training) on the trained limbs (arms or legs). In addition, it seems that transfer effect could enhance oxygen uptake at maximal exercise but not at LT or VT. However, to our knowledge, no data are available concerning transfer after both leg and arm training and at both maximal exercise and VT. Therefore, in this study the Square Wave Endurance Exercise Test (SWEET) was used to perform training with arms or legs on two groups with matched initial performance. This training simulated a 45 min session of interval training. The initial intensity was established from the individual responses to a graded exercise. Training intensity was adjusted during the program in order to achieve almost 90 % of the maximal heart rate at the end of each session. Specific and transfer effects of SWEET program were evaluated on the two experimental groups at both maximal exercise and VT level. This study could also help in understanding the determining factors of transfer effect.

Protocol

Methods

Subjects Fifteen male volunteers (age = 23 7 years [mean SD], weight = 74 12 kg and height 177 7 cm) participated in this study. Subjects were informed about the experimental procedure, and signed an informed consent. All subjects were physically active but not under specific upper or lower limb training.

Material Test sessions were performed on both wheelchair and a cycle ergometer. The wheelchair ergometer consisted of chair wheels connected to a standard friction-loaded Fleisch ergometer by means of chain and chainrings system [13, 28]. In addition, the rotation of the wheels was damped by a flywheel in order to reproduce the inertial effects encountered in track conditions. Mechanical calibration of the wheelchair ergometer was performed by comparing the power delivered by a controlled electric motor (Medifit system) and the power measured by the ergometer [28]. A significant and linear correlation (r2 = 0.997) was obtained between the reference and the measured power. Detailed description and calibration of the ergometer have previously been presented [28]. A standard magnetic-loaded cycle ergometer (1000S-MFE Medifit, MAARN, The Netherlands) was used for leg testing and training. It was calibrated before the experiment in two different conditions, unloaded and using a known mass of 50 Newton. During the tests, the subjects were connected to a gas analyser system (Oxycon Champion, Mijnhardt, The Netherlands), via a respiratory valve. The gas analyser system was calibrated be-

Before and after 6 weeks of a training program, all subjects were evaluated on two separate days on the wheelchair and the cycle ergometer. The wheelchair ergometer test consisted of 6 min rest on the wheelchair ergometer, followed by successive 2 min bouts. The exercise started at a residual friction power of about 8 watts as previously described [10 ± 13, 28]. Then the load was increased by 10 watts every 2 min until exhaustion. All exercise bouts were performed at a constant speed of 30 rpm of the wheels. The highest load, which could be performed for 2 min was taken as the peak power output Ç p). The leg exercise test was performed on the cycle erg(W ometer after a resting period of 6 min. The test started with a 3 min unloaded exercise bout [10]. The load was then inÇ p creased by 30 watts every 3 min until exhaustion. The W corresponded to the highest load that could be maintained for 3 min. In both of the tests, peak oxygen uptake (VÇO2p) and the corresponding cardiorespiratory variables (VÇEp, RERP, HRp, O2pp) Ç p. It was checked that respiratory ratio were determined at W exchange (RER) was above 1.1 or that HR was maximal (i. e., HR [220 ± age]) and HR (0.8 220 ± age) in the leg and the arm tests, respectively. The ventilatory threshold (VT) was assessed from the relation between mechanical power and selected respiratory parameters (VÇCO2/VÇO2 ratio, VÇE, VÇE/VÇO2 ratio, VÇE/VÇCO2 ratio) [39] by three blinded experts. The mean of the two closest values (out of three) was taken as the ventilatory threshold. The power Ç VT), the cardiorespiratory variables (VÇO2VT, VÇEVT, output (W Ç VT/W Ç p) and oxygen RERVT, HRVT, O2pVT), the relative power (W uptake values (VÇO2VT/VÇO2p) corresponding to the ventilatory threshold level were calculated. After the first test, the subjects were divided into three groups with matched, physical characteristics and initial performances (VÇO2p and VÇO2VT) : control group (CG, n = 5), arm-trained group (AG, n = 5) and leg-trained group (LG, n = 5). AG and LG training consisted of three SWEET sessions per week performed during 6 weeks. A session was constituted of 9 consecutive periods of 5-min including 4-min ªbaseº work Ç VT followed by 1-min ªpeakº work initially performed at the W Ç p levels, respectively. During the training period, the inand W Ç VT and W Ç p were readjusted by step tensities corresponding to W of + 10 watts for arm training and + 30 watts for leg training when the HR registered at the end of the session was at least 10 beats min±1 lower than the highest HR of the other previous sessions. Each training was then performed at the maximum intensity [10].

Training and Transfer Effects


During the study, all subjects performed usual daily activities. Control subjects did not follow training program and the trained subjects did not perform strenuous exercise with the upper or lower limbs outside their training program. Finally, all subjects were tested again after the training period using the same protocol used in the pre-training tests.

Statistical analysis All values are given as the mean and standard deviation and are expressed by the changes (in percent) obtained after training. The delta percent (D %) changes are calculated as the mean of the individual variation (value after training ± value before training/value before training 100). Training effects were evaluated by comparison of the mean D% of the changes between trained and control groups using Mann-Whitney test. For each group, mean values obtained before and after training were compared using the Wilcoxon test. The level of significance was set at 0.05 for all tests.

Results Tables 1 and 2. show the values of physiological responses at Ç p and at W Ç VT for the three groups. As expected, in pre-trainW ing tests there were no significant differences between groups Ç p or for any mechanical or physiological variable either at W Ç WVT level. In the control group, no difference was observed in the values measured in the pre and the post test performed either with arms or legs. The following section presents the comparison of D% changes between trained and control groups when the test was performed first with the trained limbs and then with the untrained limbs. Table 1

Specific effects at peak exercise Physiological values obtained by each group before and after training are presented in Table 1. After the arm test, AG showed Ç p (0 10 % vs. 66 15 %, for CG higher D% changes (p < 0.01) for W and AG, respectively), VÇO2p (1 5 % vs. 35 15 %), VÇEp (±11 11 % vs. 50 20 %) and O2pp (±2 7 % vs. 23 8 %). While peak heart rate (HRp) values were not significantly different (3 7 % vs. Ç p (2 5 % vs. 9 10 %). LG showed higher D% changes for W 14 5 %, p < 0.05 for CG an LG, respectively), VÇO2p (±1 4 % vs. 13 6 %, p < 0.01), VÇEp (3 7 % vs. 23 11 %, p < 0.01), calculated after the cycle ergometer leg test. At the same time the D% changes were not significantly different between CG and LG for the specific HRp (±1 2 % vs. ±4 5.5 %) and O2pp (0 5 % vs.11 8 %).

Specific effects at ventilatory threshold Results are presented in Table 2. Compared to CG values, AG Ç VT (14 15 % vs. 144 76 %, showed higher D% changes for W p < 0.01 for CG an AG, respectively), VÇO2VT (9 18 % vs. 51 23 %, p < 0.05), VEVT (14 9 % vs. 51 23 %, p < 0.05), O2pVT Ç VT/W Ç p (14 14 % vs. (2 18 % vs 31 12 %, p < 0.05) and W 45 33 %, p < 0.05) calculated after the arm test. While the D% changes were not significantly different between the two groups for VÇO2vT/VÇO2p (9 24 % vs 12 16 %) and HRVT (8 13 % vs 15 16 %). Compared to CG value, leg training induced highÇ VT (5 18 % vs. 53 26 %, p < 0.01 for CG and er D% changes for W LG, respectively), VÇO2VT (3 14 % vs 46 21 %, p < 0.05), VÇEVT (6 16 % vs. 78 43 %, p < 0.05), O2pVT (±1 13 % vs. 26 20 %, p < 0.05), HRVT (2 18 % vs. 30 21 %, p < 0.05). Comparison of D% changes between CG and LG also revealed higher values Ç VT/W Ç p (±3 12 % vs. 27 15 % (p < 0.05) of VÇO2VT/VÇO2p and W and 2 15 % vs. 30 17 %, respectively) in the trained group.

Effects of arm and leg training on the physiological responses at the peak level exercise during arm and leg tests (mean SD) Arm Test

Pre

Post

66 11.4***

108 10.9

228 45.5

228 45.5

30.4 4.8**

39.3 5.5

46 5.6

47 6.4

106.3 21.7

117 28.1

123 35.4

Variables

Pre

AG (n = 5)

Çp W Ç VO2p (ml kg±1 min±1) Ç Ep ( l min±1) V

72 16.6**

LG (n = 5)

CG (n = 5)

Leg Test Post

Group

RERp

0.96 0.02

0.94 0.03

1.1 0.03

1.1 0.04

HRp (bpm)

168 15.2

183 17.1

196 9.4

189 11

O2pp (ml beat±1) Çp W

12 1.8*

15 2.3

16 2.7

17 3.6

72 10.9*

80 12.2

252 34.2*

294 32.9

ÇO2p (ml kg±1 min±1) V Ç Ep (l min±1) V

30.4 2.7

34.4 1.9

46 2.6*

53 5

75.6 7.2

87.9 10.9

106 20.2*

138 10.4

RERp

0.97 0.01

0.94 0.05

1.07 0.01

1.08 0.02

HRp (bpm)

164 14.7

161 17.9

185 7.1

187 13.3

O2pp (ml beat±1) Çp W

14 2.4*

16 2.2

19 2.5*

22 3.6

60 12.2

60 14.1

240 21.2

246 25.1

ÇO2p (ml kg±1 min±1) V Ç Ep (l min±1) V

26.9 5.2

26.9 4.9

46 7.1

46 7.2

61 18.2

53.6 16.1

102 18.9

106 21.1

RERp

0.95 0.01

0.94 0.05

1.07 0.02

1.07 0.07

HRp (bpm)

159 14.6

163 14.9

191 7.4

190 6.2

O2pp (ml beat±1)

13 3

12 2.9

18 2.3

18 2.2

*, **, *** significance of the difference between the results obtained at the pre-training (PRE) and post-training (POST) leg or arm test at the level of 0.05, 0.01, 0.001, respectively

519

520


Tordi N et al

Transfer effects at peak exercise: Either with AG or LG compared to CG, no significant transfer from the trained limbs to the untrained limbs were observed Ç p and HRp. LG showed higher D% changes for VÇEp for VÇO2p, W and O2pp in the arm test, compared to the CG values (-11 11 % vs. 17 17 %, p < 0.05 and -2 7 % vs. 16 9 %, p < 0.05, respectively). AG showed higher D% changes for O2pp in leg test, compared to the CG values (0 5 % vs. 6 6 %, p < 0.05).

Transfer effects at ventilatory threshold: For AG, the D% changes observed during leg test were significantly different from those observed in CG, for VÇO2VT/VÇO2p Ç VT (±3 12 % vs. 19 9 %, p < 0.01 for CG an AG, respectively), W (±1 13 % vs. 19 14 %, p < 0.01) and VÇO2VT (±3 14 % vs. 23 14 %, p < 0.01). No differences were noted with the other physiological parameters. During arm test, leg training induced significant changes for VÇO2VT (9 18 % vs. 46 23 %, p < 0.05 for CG an LG, respectively), O2PVT (2 18 % vs. 30 21 %, p < 0.05) and VÇEVT (14 9.4 % vs. 64 46 %, p < 0.05). Ç VT was not significantly different However, D% changes for W (15 14 % vs. 33 6 %) between the two groups.

Discussion The significant changes observed with the trained groups may be related to the effects of training. Indeed none of the modifications observed with the trained groups could be observed

Table 2

with the CG. However, it can be argued that for some variables the changes did not reach the significance level because of the small number of subjects involved in the study.

Specific effects at peak exercise For the three groups, the initial levels obtained with the arms or the legs are in agreement with those of previous studies performed with a wheelchair [28] or bicycle ergometer [10]. In AG, VÇO2p increase (+ 35 %, Table 1) was about twice than usually found in the literature with other types of training (Table 3). For instance, Bhambhani et al. [3] and Magel et al. [22] have reported a 17 % and 16 % increase, respectively. Only Lewis et al. [20] obtained comparable improvement (35 %) probably because sedentary subjects with low peak oxygen uptake before training were studied. This result shows that, when the initial level of subjects is taken into account, the SWEET training is one of the best programs which can increase VÇO2 during arm-trained exercise. In LG, VÇO2p increase (+ 14 %, Table 1) was similar to the results obtained by Rosler et al. [31] (+ 13 %) and Lewis et al. [20] (+ 15 %). Other studies have found lower values (+ 8 %) [3] or higher values (+ 30 %) [10]. However, in the study of Gimenez et al. [10], the sedentary status of subjects could explain the high training effect. Therefore, it seems that SWEET training is also well adapted to optimize VÇO2 increase during leg exercise.

Effects of arm and leg training on the physiological responses at the ventilatory threshold (VT) during arm and leg tests (mean SD) Arm Test

Leg Test

Group

Variables

Pre

Post

Pre

Post

AG (n = 5)

Ç VT W Ç VT/Wp (%) W Ç VO2VT (ml kg±1 min±1)

32 11*

72 8.4

132 34.2*

156 39.1

LG (n = 5)

47 10*

66 3

57 8*

67 6

17 2.5*

25.4 3.4

28 6.2*

34 5.8

ÇO2p (%) ÇO2VT/V V Ç VEVT (l min±1)

57 4.3

64 4.6

59 7*

68 5.8

31 5.5*

48 11.9

46 11.9

57 19

HRVT (bpm)

127 12.6*

146 19.7

148 20.6

153 17

O2pVT (ml beat±1) Ç VT W

9 1.9*

12 2.1

13 2.5

15 2.9

36 8.9

48 13

138 40.2*

204 32.9

49 7*

57 8

53 9*

67 4.5

Ç VT/Wp (%) W ÇO2VT (ml kg±1 min±1) V

CG (n = 5)

15 2.6*

21.5 2.4

26 5.2*

37 4.6

ÇO2VT/VO2p (%) V Ç VEVT (l min±1)

49 6.3*

63 8.8

56 8.4*

70 4

30 4.9*

48 7.5

42 8.9*

72 10.8

HRVT (bpm)

116 10.7

131 14

135 11.9*

157 12.1

O2pVT (ml beat±1) Ç VT W

10 2.3*

13 1.5

15 2.6

18 3.4

34 11.4

38 14.8

138 16.4

144 17.3

57 10

58 12

57 5

58 6.2

Ç VT/W Ç p (%) W ÇO2VT (ml kg±1 min±1) V

17 5.4

18.1 4.2

29 6.3

28 3.4

ÇO2VT/V ÇO2p (%) V Ç EVT (l min±1) V

62 10

66 5

62 5.7

60 3.6

30 8.6

34 8.2

44 6.4

47 11.2

HRVT (bpm)

126 20.6

136 26.9

158 16.7

154 7.3

O2pVT (ml beat±1)

10 2.4

10 3.2

13 1.7

13 2.3

*, **, ***significance of the difference between the results obtained at the pre-training (PRE) and post-training (POST) leg or arm test at the level of 0.05, 0.01, 0.001, respectively.



The significant VÇO2p increase obtained in arm exercise and, with less extension, in leg exercise could be explained by the higher duration and intensity level of SWEET training compared to training programs usually recommended for healthy subjects [2]. These recommendations usually include 15 to 60 min of rhythmic exercise, 3 to 5 days per week, at 50 % to 85 % of VÇO2max or 60 % to 90 % of maximum heart rate reserve. The high intensity level during SWEET training could be obtained Ç p and W Ç VT by adjustment of training from individual data at W levels, in such a way that it optimizes the use of aerobic metabolism and it avoids lactate accumulation [11] during 45 min.

Specific effects at ventilatory threshold For the AG, the levels of pre and post training VT correspond, respectively, to those of untrained and trained paraplegic subjects [5]. This confirms the effectiveness of training and method used to determine the VT levels. To the best of our knowledge, only the study of Bhambhani et al. [3] analysed both the specific and transfer effects obtained at VT. The present study showed higher VÇO2VT improvements than those from the Bhambhanis study during both the AG arm (+ 51 % vs. + 23 %) Table 3

and LG leg test (+ 53 % vs. + 13 %). These discrepancies could be attributed to the specific metabolic responses induced by the SWEET. During continuous exercise, lactate accumulation occurring for exercise intensity above anaerobic threshold, results in metabolic acidosis that cannot be compensated by physiological mechanisms [17]. On the contrary, during the SWEET, compensation mechanisms could take place. Hyperventilation produces ventilatory alkalosis during and after the peak level and, by glyconeogenesis, allows the muscles to transform lactate into glucose during the base level [11]. SWEET training also induced very high VÇEVT increases (+ 51 % and + 78 % in arm and leg test) compared to continuous training (+ 9 and + 12 % in arm and leg test) [3], further demonstrates the role of hyperventilation in the reduction of lactate accumulation during the SWEET exercise. It should also be noted that VÇO2VT/VÇO2p ratio was significantly increased in LG (+ 25 %) while it was not significantly changed in AG (+ 11 %, p = 0.06). These discrepancies between arm and leg results may be related to different phenomenon. Firstly, Magel et al. [22] and Loftin et al. [21] suggested that the arm muscles have a relatively larger potential than leg muscles for local metabolic and circulatory improvement through training. Secondly,

ÇO2 and the ventilatory threshold Summary of studies on the arm and leg training and the transfer effects on the peak V

Study

Subjects

Training mode

Results of testing (VO2)

Intensity

Length

Duration (min)

Peak

D%

VT

D%

5W

I-maxHR

3d wk±1 6-wk

45

W = 30.4 L = 46

+ 35** +2

17 28

+ 51* + 22*

5L

I-maxHR

3d wk±1 6-wk

45

L = 46 W = 30.4

+ 14* + 13

26 15

+ 53* + 46*

8A

C-72 % A-VO2p

3d wk±1 8-wk

30

A = 32 L = 44.5

+ 17* +2

13.8 27.5

+ 23* +1

8L

C-80 % L-VO2p

3d wk±1 8-wk

30

L = 43.8 A = 28.9

+ 9* +1

25.7 14

+ 13* +1

Loftin et al. 1988

19A

I-70 ± 90 % HR reserve

4d wk±1 5-wk

64

A=± L= ±

+ 32* + 7*

± ±

± ±

Rosler et al. 1985

10L

C-90 ± 95 % maxHR

5d wk±1 8-wk

30

L = 51.6 A = 38.4

+ 13* + 9*

± ±

+ 27*a

Lewis et al. 1980

5L

C- ³ 150 170bmin±1

3d wk±1 10-wk

30

L = 39.2 A = 25

+ 15* + 9*

± ±

± ±

5A

C- ³ 150 170 b min±1

3d wk±1 10-wk

30

A = 22.8 L = 37.2

+ 35* + 12*

± ±

± ±

8A

C- ³ 180 b min±1

3d wk±1 10-wk

10

A = 36.9 BS = 42.7

+ 19* +1

± ±

± ±

9BS

C- 180 ± 190 b min±1

3d wk±1 10-wk

15

BS = 42.1 A = 37

+ 15* 0

± ±

± ±

Magel et al. 1978

7A

I- ³ 85 % maxHR

3d wk±1 10-wk

64

A = 33.9 T = 56.4

+ 16* +1

± ±

± ±

Ridge et al. 1976

5K 5L

C- 85 ± 90 % maxHR

4d wk±1 4-wk

30

K = 31.9 K = 30.7

+ 8* +1

± ±

± ±

Pollock et al. 1974

8A

C- 150 ± 165 b min±1

3d wk±1 20-wk

30

A = 23.3 T = 37.9

+ 39* + 7*

± ±

± ±

Clausen et al. 1973

3L

I-170 b min±1

5d wk±1 5-wk

45

L= ± A=±

+ 17* + 10*

± ±

± ±

Present

Bhambhani et al.1991

Stamford et al.1978

* p < 0.05, ** p < 0.01 Abbreviations: d = days, wk = weeks, C = Constant exercise, I = interval exercise, L = leg cycling, A = arm cycling, T = treadmill running, W = wheelchair, BS = bench ÇO2 = oxygen uptake ml kg±1 min±1, VT = initial ventilatory threshold, a = increase of the power output at the 4 mmol l±1 lactate level, stepping, K = kayak, V D% mean of the individual percent change

521

522


the initial level of fitness is a major factor for improvement in exercise capacity [34, 40] and it is accepted that the arm muscles are less requested to heavy rhythmic oxidative exercise than leg muscles in the daily activities.

Transfer effects at peak exercise Ç p (+ 11 %) in arm In this study, LG significantly improved W tests, while no improvement was found in leg tests performed by AG. A review on this topic [9] indicated that the results of the transfer effect of training are to date inconclusive. After leg training, Lewis et al. [20] and Rösler et al. [31] have reported significant improvements (9 %) in peak oxygen uptake during arm testing. After arm training, Lewis et al. [20], Loftin et al. [21] and Pollock et al. [25] have reported significant improvements (respectively, 12 %, 7 % and 7 %) in peak oxygen uptake, during leg testing. However, other authors [3, 22, 23, 30, 36] have suggested that improvements are specific for the trained muscle group. Even if conditions, under which transfer effects of training may occur, are controversial, it was hypothesized that peripheral [31] or central adaptations [4, 20, 21], induced by endurance training, might be responsible for the transfer effects. Peripheral effects of endurance training include aerobic adaptations of muscle tissue, such as an increased capillarization [18], an increased in the mitochondrial enzyme activity [14], and/or an increased in the mitochondrial volume density [18]. However, Rosler et al. [31] showed that leg endurance training has no effect on the ultrastructure of untrained muscle tissue of the deltoideus. Although we have no direct evidence that the ultrastructure of the muscle changed in our subjects, it seems very unlikely that such phenomenon could explain the transfer effect observed in the present study. It could also be argued that static contractions of untrained limbs, occurring during training of other limbs, might be responsible for ªtransferº effect. However, the presence or lack of transfer effects to untrained limbs does not appear to depend on whether stabilization work with untrained limbs has [3] or has not been [36] permitted during the testing and training procedures. Furthermore, static or heavy resistance dynamic training has essentially no effect on the cardiovascular responses and endurance of dynamic arm or leg exercise [1]. Therefore, the transfer effect observed in the present study was probably not due to the muscular contractions of the untrained limbs during arm or leg training. In endurance training, central modifications are reflected by an increase of blood volume, stroke volume and cardiac output [4, 34, 38]. Therefore, the increase in VÇO2p during work with untrained limbs could be attributed to central adaptations, which allows a better oxygen uptake by the untrained muscles [21, 38]. After arm or leg training, Thompson et al. [38] realized an echocardiographic evaluation of cardiac performance during exercise involving trained or untrained muscles. During both types of exercises, their results suggested an increase in the stroke volume subsequent to training. The significant increase of O2pp (+ 16 %) observed LG in the arm test, further supports this explanation. In addition, the transfer effect of training on the peak oxygen uptake was only observed from the legs to the arms. This could also be due to the lower central adaptations in arm training compared to leg training [4, 38]. In the arm training, due to the lower muscle mass involved, the max-

Tordi N et al

imal heart rate was equivalent to 85 ± 90 % of maximal heart rate level of leg exercise. As a consequence, the cardiovascular system was not fully taxed during the arm training program and central adaptation was not sufficient to induce transfer effects after arm training. On the contrary, the leg training fully taxed the cardiovascular system, allowing a transfer effect to the arm. Therefore, it seems that central adaptations were the main factors responsible for transfer effect measured at peak exercise. Moreover, Swensen et al. [37] suggested that the amount of muscles involved in the training could also influence the transfer effect at peak exercise. This could explain why in the present study the transfer phenomenon seems to be more effective from leg to arm than the other way around.

Transfer effects at ventilatory threshold An original finding of the present study was the observation of transfer effects measured on the VÇ02VT after both arm and leg training (+ 46 % and + 22 %, respectively). These results are not in line with the literature. For instance, Bhambhani et al. [3], did not show any transfer effect on VÇO2VT after constant training with arm or leg and Rösler et al. [31] did not observe any changes in the power output at the 4 mmol l-1 lactate level measured during the test with the untrained arms. This discrepancy between the results of the present study and of those from the literature is probably due to the high intensity and the specificity of SWEET training. After low intensity training, lactate production and concentration are decreased only at low intensity exercise, whereas it has been suggested that, after high intensity training, the lactate clearance is increased and blood lactate concentration is reduced at high intensity exercise [24]. Although lactate uptake from blood can occur in many tissues e. g., heart and liver, it has been shown that during exercise skeletal muscle probably represents the major site of blood lactate clearance. The lactate produced by recruitment of type IIb fibres is transported into type I and IIa fibres, where it is oxidized. By this way trained but non-exercising muscles could be able to take up lactate by exercising muscles [15,19, 27, 31, 41]. As a consequence, lactate decreased and lactic acidosis is diminished if not entirely compensated [11]. Therefore, although we did not measure lactate uptake capacity induced by training, it could be hypothesized that the transfer effect, observed in the present study at VT level, was a consequence of a higher lactate uptake and oxidation by the trained but non-exercising muscles. It is also worth noting that after training LG revealed a significant increase of VÇO2vT/VÇO2p (from 0.49 to 0.63, + 28 %) during the arm test and that arm training induced the same kind of VÇO2VT/VÇO2p changes (from 0.59 7 to 0.68 5.8, + 15 %) during the cycle ergometer test. This result shows that SWEET training was more effective at ventilatory threshold than at peak intensity exercise, especially for the AG. Poole and Gaesser [26] have previously compared the effect of leg training at constant workload (34 min at 70 % VÇO2max) with that of interval leg training (10 cycles of 2 min 105 % VÇO2max and 2 min rest). They found that the interval training improves the VÇO2VT more than the constant training. In addition, Gimenez et al. [10,11] showed that SWEET conducted with the legs induced maximal heart rate, and normal values [H+] and lactate concentration at least 5 mmol l±1. This relative homeostasis appeared to arise


from periods of hyperventilation observed during and after each ªpeakº and the consumption of lactate by the muscles during work at the ªbaseº [12]. Therefore, the SWEET seems to be appropriate to induce maximal transfer effect at VT level.

Conclusion In the present study, both specific and transfer effects obtained after leg or arm SWEET training have been observed. Compared to other types of training already tested in the literature, and taking into account initial level of the subjects and intensity level of training, we are able to show that SWEET training induces higher specific and transfer effects than other types of training. At maximal level, transfer effects of training were more efficient from the legs to the arms than from arms to the legs, probably due to peripheral limitations occurring during arm exercise where central cardiorespiratory adaptations are low. At the VT level, SWEET training also induced significant transfer effects in both arm and leg exercises. These findings were consistent with the greater capacity for lactate metabolism resulting from endurance training. This type of training program could then be very useful for either the rehabilitation or the training of paraplegic subjects.

Acknowledgements The authors thank the subjects for their enthusiastic and regular participation, and the technical staff of the UnitØ 14 INSERM (C.O. 10, 54 511 Vandoeuvre ls Nancy, France) for their assistance. Benoit Duguet Ph. D. is thanked for his help in finalizing the manuscript.

References 1

Allen TE, Byrd RJ, Smith DP. Hemodynamic consequences of circuit weight training. Res Quart 1978; 47: 299 ± 306 2 American College of Sports Medicine. The recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness in healthy adults. Med Sci Sports Exerc 1990; 22: 265 ± 274 3 Bhambhani YN, Eriksson P, Gomes PS. Transfer effect of endurance training with the arm and leg. Med Sci Sports Exerc 1991; 23: 1035 ± 1041 4 Clausen J, Klausen K, Rasmussen B, Trap-Jensen J. Central and peripheral circulatory changes after training of the arm or leg. Am J Physiol 1973; 225: 675 ± 682 5 Coutts KD, McKenzie DC. Ventilatory threshold during wheelchair exercise in individuals with spinal cord injuries. Paraplegia 1995; 33: 419 ± 422 6 Davis JA, Frank M, Whipp B, Wasserman K. Anaerobic threshold alterations caused by endurance training in middle age men. J Appl Physiol 1979; 46: 1039 ± 1046 7 Davies CTM, Sargeant AJ. Effects of training on the physiological responses to one or two leg work. J Appl Physiol 1975; 38: 377 ± 381 8 Denis C, Fouquet R, Poty P, Geyssant A, Lacour JR. Effect of 40 weeks endurance training on anaerobic threshold. Int J Sports Med 1982; 3: 208 ± 214 9 Franklin BA. Aerobic exercise training programs for the upper body. Med Sci Sport Exerc 1989; 21: 141 ± 148 10 Gimenez M, Servera E, Salinas W. Square-Wave Endurance Exercise Test (SWEET) for training and assessment in trained and untrained subjects. I) Description and cardiorespiratory responses. Eur J Appl Physiol 1982; 49: 359 ± 368

Int J Sports Med 2001; 22 11

Gimenez M, Servera E, Saunier C, Lacoste J. Square Wave Endurance Exercise Test (SWEET) for training and assessment in trained and untrained subjects. II) Blood gases and acid-base balance. Eur J Appl Physiol 1982; 49: 369 ± 377 12 Gimenez M, Cereceda V, Teculescu D, Aug F, Laxenaire MC. Square Wave Endurance Exercise Test (SWEET) for training and assessment in trained and untrained subjects. III) Effect on VO2max and maximal ventilation. Eur J Appl Physiol 1982; 49: 379 ± 387 13 Gimenez M, Predine E, Marchand M, Kittstein G. Validation dun fauteuil-roulant ergomtre pour la mesure de la rØponse cardiorespiratoire et le rendement mecanique. Arch Intern Physiol Bioch 1991; 99: 102 14 Gollnick PD, Saltin B. Significance of skeletal muscle oxidative enzyme enhancement with endurance training. Clin Physiol 1982; 2: 1 ± 12 15 Granier P, Dubouchaud H, Mercier B, Mercier J, Ahmaidi S, Prefaut C. Lactate uptake by forearm skeletal muscles during repeated periods of short-term intense leg exercise in humans. Eur J Appl Physiol 1996; 72: 209 ± 214 16 Henritze J, Weltman A, Schurrer L, Barlow K. Effects of training at and above lactate threshold on lactate threshold and maximal oxygen uptake. Eur J Appl Physiol 1985; 54: 84 ± 88 17 Hermansen L, Stensvold I. Production and removal of lactate during exercise in man. Acta Physiol Scand 1972; 86: 191 ± 201 18 Hoppeler H, Howald H, Conley K. Endurance training in humans: Aerobic capacity and structure of skeletal muscle. J Appl Physiol 1985; 59: 320 ± 327 19 Klausen K, Secher NS, Clausen JP, Hartlin GO, Trap-Jensen J. Central and regional circulatory adaptations to one leg training. J Appl Physiol 1982; 52: 976 ± 983 20 Lewis S, Thompson P, Areskog NH. Transfer effects of endurance training to exercises with untrained limbs. Eur J Appl Physiol 1980; 44: 25 ± 34 21 Loftin M, Boileau A, Massey BJ, Lohman TG. Effect of arm training on central and peripheral circulatory function. Med Sci Sports Exerc 1988; 20: 136 ± 141 22 Magel JR, Mcardel WD, Michael T, Delio DJ. Metabolic and cardiovascular adjustment to arm training. J Appl Physiol 1978; 45: 75 ± 79 23 Mckenzie DC, Fox EL, Cohen K. Specificity of metabolic and circulatory responses to arm or leg interval training. Eur J Appl Physiol 1978; 39: 241 ± 248 24 McRae HSH, Dennis SC, Bosch AN, Noakes TD. Effect of training on lactate production and removal during progressive exercise in humans. J Appl Physiol 1992; 72: 1649 ± 1656 25 Pollock J, Miller H, Linnerud A, Laughridge E, Coleman E, Alexander E. Arm pedalling as an endurance training regimen for the disabled. Arch Phys Med Rehab 1974; 55: 418 ± 424 26 Poole DC, Gaesser GA. Response of ventilatory and lactate thresholds to continuous and interval training. J Appl Physiol 1985; 58: 1115 ± 1121 27 Poortmans JR, Delescaille-Vanden Bossche J, Leclercq R. Lactate uptake by inactive forearm during progressive leg exercise. J Appl Physiol 1978; 45: 835 ± 839 28 PrØdine E, Gimenez M. Mise au point et validation dun prototype de fauteuil roulant ergomtre permettant la mesure defforts rØalisØs avec les membres supØrieurs. Medecine du Sport 1996; 70: 17 ± 24 29 Rasmussen B, Klausen K, Clausen JP, Trap-Jensen J. Pulmonary ventilation, blood gases, and blood pH after training of the arm or the leg. J Appl Physiol 1975; 38: 250 ± 256 30 Ridge BR, Pyke FS, Roberts AD. Responses to kayak ergometer performance after kayak and bicycle ergometer training. Med Sci Sports Exerc 1976; 8: 18 ± 22 31 Rösler K, Hoppeler H, Conley KE, Claassen H, Gehr P, Howald H. Transfer effect in endurance exercise: adaptations in trained and untrained muscles. Eur J Appl Physiol 1985; 54: 355 ± 362

523

524

Int J Sports Med 2001; 22 32

Rusko H, Rahkila P, Karvinen E. Anaerobic threshold, skeletal muscles enzymes and fiber composition in young female crosscountry skiers. Acta Physiol 1980; 108: 263 ± 268 33 Saltin B, Hartley LH, Kilbom A, Astrand I. Physical training in sedentary men. Scand J Clin Lab Invest 1969; 24: 323 ± 324 34 Saltin B, Nazar K, Costill DL. The nature of the training response; peripheral and central adaptations to one legged exercise. Acta Physiol Scand 1976; 96: 289 ± 305 35 Seilvert GG, Wenger HA. Physiological predictors of short-course triathlon performance. Med Sci Sports Exerc 1993; 25: 871 ± 876 36 Stamford BA, Cudihee RW, Moffatt RJ, Rowland R. Task specific changes in maximal oxygen uptake resulting from arm versus leg training. Ergonomics 1978; 19: 53 ± 60 37 Swensen TC, Howley ET. Effect of one- and two-leg training on arm and two-leg maximum aerobic power. Eur J Appl Physiol 1993; 66: 285 ± 288 38 Thompson SD, Lewis S, Varady A. Cardiac dimensions and performance after either arm or leg endurance training. Med Sci Sports Exerc 1981; 13: 303 ± 309 39 Wasserman K. The anaerobic threshold: definition physiological significance, and identification. Adv Cardiol 1986; 35: 1 ± 23 40 Wenger H, Bell G. The interactions of intensity, frequency and duration of exercise training altering cardiorespiratory fitness. Sports Medicine 1986; 3: 346 ± 356 41 Yoshida T, Takeuchi N, Suda Y. Arterial venous blood lactate increase in the forearm during incremental bicycle exercise. Eur J Appl Physiol 1982; 50: 887 ± 893

Tordi N et al

Corresponding Author: N. Tordi, Ph. D. Laboratoire des Sciences du Sport Place Saint Jacques 25030 BesancËon cedex France Phone: Fax: E-mail:

+33 (3) 8166-5617 +33 (3) 8166-5692 [email protected]